Mining Comprehensible Classification Rules with Metaheuristic Strategy: AntMiner

In mining, classification rules from knowledge-based discovery and expert systems. Nature-inspired algorithms are gaining popularity these days due to their metaheuristic approach used in discovering classification models in the sphere of data-mining. This paper explores an experimental study on six datasets by implementing AntMiner. Simulation study is carried out by comparing AntMiner with existing rule creation algorithmic strategies (JRIP and PART). NP-test are conducted for evaluating the performance measures of various algorithms in mining rules of classifiers and leads to the fact that cAntMiner-2 as well as cAntminer-PBoutperform over JRIP and PART algorithms.